When communicating over circuit-switched networks, such as Public Switched Telephone Networks (PSTNs) or legacy Private Branch Exchanges (PBXs), most voice data transmissions reflect some signal energy back towards their source. This phenomenon, commonly known as echo or hybrid echo, typically occurs at physical transitions within the circuit-switched networks, e.g., 4-wire to 2-wire conversion points. Since perception of reflected signal energy by callers can dramatically degrade the quality of a call, modern communication systems include one or more echo cancellers to detect and cancel hybrid echo prior to caller perception.
Echo cancellers typically cancel hybrid echo according to a plurality of echo cancellation parameters, i.e., input gain, output attenuation, impedance, echo return loss (ERL), combined signal loss (ACOM), etc. Between call termination and establishment of the next call, these echo cancellation parameters are often reset to a default setting and then automatically tuned according to the hybrid echo generated during each call. Although this automatic tuning enables echo cancellers to dynamically eliminate hybrid echo associated with the calls, a period of time exists prior to the automatic tuning where the echo cancellers rely upon the default setting to cancel the hybrid echo. Since the default setting of the echo cancellation parameters commonly requires significant adjustment before the echo cancellers can effectively eliminate the hybrid echo, some hybrid echo may be perceived by callers prior to the completion of the automatic tuning, thus degrading call quality.